Fig 1.
Detection of AIGA and nAIGA levels in plasma samples.
(A) Levels of AIGAs in plasma from AOID patients (n = 40) and healthy individuals (n = 40) were measured using indirect ELISA. Cut-off value for indirect ELISA was defined as the mean OD₄₅₀ of HC samples plus two standard deviations. (B) The nAIGA levels in these samples were assessed using a cell-based functional assay in THP-1 cells. Cut-off value for this assay was defined as the mean % inhibition of HC samples plus two standard deviations. HC = HC samples; AIGA+ = AIGA-positive samples. The statistical difference between two groups was determined by Mann-Whitney U test (**** p < 0.0001).
Fig 2.
Optimization of cELISA for detecting nAIGAs.
(A) Schematic representation of the cELISA principle for detecting B27 epitope‑targeting nAIGAs in plasma. Plasma samples were preincubated with B27 mAb-HRP and subsequently applied to microwells coated with rhIFN‑γ. In the absence of B27 epitope‑targeting nAIGAs, B27 mAb-HRP binding to rhIFN‑γ was unimpeded, producing a strong signal (upper panel). Conversely, in the presence of B27 epitope‑targeting nAIGAs, competitive inhibition prevented B27 mAb binding to rhIFN‑γ, resulting in diminished signal following substrate addition (lower panel). (B) Heatmap of OD₄₅₀ signals generated by serial dilutions of B27 mAb-HRP binding to microwells coated with increasing concentrations of rhIFN‑γ. The mean OD₄₅₀ values of duplicate wells from two independent experiments were used to construct the heatmap. (C) Validation of the optimized condition using non‑labeled B27 mAb (0–10 µg/mL) competing with B27 mAb-HRP (1:2,500 dilution). The bar graph shows the mean ± SD from duplicate wells.
Fig 3.
Detection of B27 epitope-targeting nAIGAs using cELISA.
The inhibition levels of B27 epitope-targeting nAIGAs in plasma samples from HCs (n = 40) and AOID patients (n = 40) were assessed using cELISA. Each sample was assayed in duplicate, and the mean OD450 was used to calculate the % inhibition. The assay cut-off was defined as 8.89% inhibition, calculated from the mean percentage inhibition observed in HC samples (−22.3%) plus two standard deviations (2SD = 31.2%). HC = healthy control samples; AIGA⁺ = AIGA-positive samples. The statistical difference between AIGA+ and HC was determined using Mann-Whitney U test. **** p ≤ 0.0001.
Fig 4.
ROC analysis of assays for detecting nAIGAs.
ROC analysis was performed to evaluate the diagnostic performance of two assays. The AUC provides a measure of diagnostic accuracy, with values between 0.800 and 1.000 considered indicative of good to excellent performance. Sensitivity and specificity were also plotted against cut-off values to determine the optimal threshold for each assay. (A) ROC curve and (B) sensitivity–specificity versus cut-off plot of the cell-based functional assay. (C) ROC curve and (D) sensitivity–specificity versus cut-off plot of cELISA. The red dashed line indicates the optimal cut-off that maximizes both sensitivity and specificity.
Fig 5.
Correlation between cELISA and the cell-based assay.
The association between percentage inhibition values obtained from the two methods was evaluated using Spearman’s correlation coefficient. A dotted line indicates zero inhibition as a reference. The correlation coefficient (r) and corresponding P-value are displayed on the plot.